本試驗試區於六十六年元月將原第一階段試驗之戀風草草帶改植百喜草,除對照區外,並在試區中央加設1公尺寬百喜草草帶,每一試區分成上,下坡小區,試區寬8公尺,長13公尺,耕作坡長上,下坡小區各爲5公尺,對照區無草帶設置,耕作坡長仍爲13公尺。以雜作不同處理之二年輪作方式從事試驗,田間設計採用完全逢機區集法,四重複,五處理,試驗處理包括:(1)密植區:高粱-大豆-玉米(第一年)落花生-甘藷(第二年),(2)春、夏作密植區:甘藷-玉米(第一年)高粱-落花生-小麥(第二年)(3)落花生-甘藷(第一年)玉米-大豆-高粱(第二年),(4)木薯(第一年)-大豆-甘藷(第二年),(5)對照區:高粱-落花生-甘藷(第一年)-大豆-玉米(第二年),茲將試驗結果摘要如下:1.以雨量分佈而言,雜作春作初、中期及秋作中、後期,均因缺乏足夠雨量,致其產量與品質俱劣,夏作則因長期雨季,高溫、多濕、寡照,作物容易徒長,收益低微。2.本試驗之百喜草草帶設置配合雜作之雨季密植栽培,有助抑制水土流失,惟作物應選擇耐高溫抗濕及較具粗放栽培而地面覆蓋良好之品種如甘藷台農五七號等。3.各期雜作之水土流失均以生育初、中期或雨季始期較爲龐大,此完全因整地、中耕、施肥等作業之施行,造成表土鬆軟助長了逕流與沖蝕4.百喜草草帶初植年份覆蓋未臻緊密,根系分佈淺窄,水土保持功能較弱,但種植滿一年後草帶結構堅實,根系分佈深廣,水土保持功能甚佳。百喜草草帶應於乾季種植,并妥予灌溉施肥以促其早日成長定型覆蓋。5.百喜草草帶二年間淤土形成堦段之程度,各小區以上邊坡田區坡度降低較小,平均約3°,下邊坡田區坡度降低幅度較大,平均爲6.66°。二年間各草帶處理試區土壤流失總量爲161 .94MT/ha-295.04MT/ha而對照區坡度降低雖衹有1.5°,但二年間土壤流失量却高達487.26 MT/ha,比草帶處理區高出39%-67%,足見百喜草草帶之處理殊有必需性。
The experimental site was established in 1977. A bahiagrass barrier (1 m wide) was planted across the slope and in the center of the plot which is 8 m wide and 13 m long. The experiment was in randomized block design Five treatments in four replications were applied: (1) close planting: sorghum-soybean-corn (first year) -peanuts-sweet potatoes (second year); (2) spring-summer crop close planting : sweet potatoes-corn (first year)-sorghum-peanuts-wheat (second year) ;(3) pea- nuts-sweet potatoes (first year)-corn-soybean sorghum (second year); (4) cassava (first year) soybean-sweet potatoes (second year); (5) control: sorghum-peanuts -sweet potatoes (first year)-soybean-corn (second year). The results obtained are summarized as follows: 1. Bahiagrass barrier in combination with close planting of upland crops that can provide good ground cover will significantly reduce soil loss. 2. Great amounts of soil and water losses occur at the earlier stage of crop growth and rainy season due to frequent cultivation practices needed 3. During the first year, the bahiagrass barrier is still not well established and thus soil and water losses are high. But during the second year, the losses are markedly reduced. 4 During the two year period of experiment, the slope of the upper portion of the plot, i.e, above the grass barrier reduce averagely by 3° while the slope of lower portion of the plot by 6.7°. The amounts of soil losses from plots with grass barrier range from 162 to 295 MT/ha in two years. For the check plot, the slope reduce by only 1.5°, but soil loss is 487 MT/ha in two years. This indicates the necessity of bahiagrass barrier for growing upland crops on sloping land.